Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry

ARTICLE IN PRESS Trends in Analytical Chemistry ■■ (2015) ■■–■■

Contents lists available at ScienceDirect

Trends in Analytical Chemistry j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / t r a c

Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry Enric Brillas, Ignasi Sirés * Laboratori d’Electroquímica dels Materials i del Medi Ambient, Departament de Química Física, Facultat de Química, Universitat de Barcelona, Martí i Franquès 1-11, 08028 Barcelona, Spain

A R T I C L E

I N F O

Keywords: Electrochemical oxidation Electro-Fenton GC-MS Hydroxyl radical LC-MS Oxidation product Pharmaceutical Photoelectro-Fenton Sunlight Wastewater treatment

A B S T R A C T

Electrochemical technology has attracted increasing interest in recent years as an environment-friendly solution to many industrial problems and challenges. First, we give an overview on the fundamentals of electrochemical processes for the removal of pharmaceuticals from water, particularly electrochemical oxidation and Fenton-based processes, such as electro-Fenton and ultraviolet (UV) and solar photoelectroFenton, that have improved performance. We also mention other less studied methods, although the main focus is on reactivity elucidation by chromatography with UV or conductivity detection, especially by mass spectrometry techniques (e.g. coupled to gas chromatography or liquid chromatography), in order to discuss the degradation pathways of pharmaceuticals on the basis of the reactive species electrogenerated in each technology. In some cases, simultaneous assessment of toxicity adds crucial information for the future integration of these technologies in water-treatment facilities where pharmaceuticals and their byproducts can occur. © 2015 Elsevier B.V. All rights reserved.

Contents 1. 2.

3.

4.

Introduction ............................................................................................................................................................................................................................................................. Electrochemical technologies for removing pharmaceuticals from water streams ........................................................................................................................ 2.1. Electrochemical oxidation ..................................................................................................................................................................................................................... 2.2. Electro-Fenton and photoelectro-Fenton ......................................................................................................................................................................................... 2.3. Other electrochemical methods .......................................................................................................................................................................................................... Elucidation of reaction products by mass spectrometry .......................................................................................................................................................................... 3.1. GC-MS ........................................................................................................................................................................................................................................................... 3.1.1. Use of HPLC, IC and toxicity analysis to clarify the reactivity of pharmaceuticals ........................................................................................... 3.2. LC-MS ........................................................................................................................................................................................................................................................... Conclusion and outlook ....................................................................................................................................................................................................................................... Acknowledgements ............................................................................................................................................................................................................................................... References ................................................................................................................................................................................................................................................................

1. Introduction Over the past 20 years, the role of pharmaceuticals as potential bioactive chemicals in the environment has received increasing attention [1]. Thousands of tons of these compounds are yearly produced worldwide for usage in human and veterinary medicines, and agricultural and consumer products. Consequently, their occurrence in the aquatic environment may arise from direct

* Corresponding author. Tel.: +34 934039240; Fax: +34 934021231. E-mail address: [email protected] (I. Sirés).

1 2 2 4 4 5 5 6 8 8 9 9

disposal, excretion and treatments in animal-feeding operations, such as aquaculture [2]. The inefficient removal of these pharmaceuticals by conventional processes used in wastewater-treatment facilities (WWTFs) causes the presence of a large plethora of drugs and metabolites in surface, ground and drinking waters at contents from nanograms to micrograms per liter (ng/L to μg/L) [1,2]. These pollutants affect the quality of water and are considered emerging pollutants, since they are unregulated, thus becoming potentially toxic for ecosystems and living beings. Even low drug contents can favor the development of multi-resistant strains of microorganisms, may affect the endocrine system of fishes and can exert toxic effects on algae and invertebrates [2,3]. Furthermore, the toxicity of complex drug cocktails, such as those routinely detected,

http://dx.doi.org/10.1016/j.trac.2015.01.013 0165-9936/© 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS 2

E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

has seldom been predicted [1]. Many research efforts are being devoted to develop efficient techniques for total removal of pharmaceuticals and their products from water. Recent reviews described effective methods, such as advanced oxidation processes (AOPs), for the remediation of water containing pharmaceuticals [1,3–7]. AOPs are considered environmentallyfriendly technologies because no hazardous additives or only small amounts of non-toxic substances are added to the solution. They are chemical, photochemical or electrochemical methods that allow the in-situ generation of the hydroxyl radical (•OH), which is the second strongest oxidizing agent known (after fluorine). Due to its high standard redox potential [E°(•OH/H2O) = 2.8 V/SHE], it is able to react nonselectively with most organics via dehydrogenation or hydroxylation steps until their total mineralization to CO2, H2O and inorganic ions [1,3]. Recently, several electrochemical AOPs (i.e. EAOPs) were developed and applied to the treatment of pharmaceuticals in water [3]. The most promising EAOPs are electrochemical oxidation (EO) and indirect methods [e.g. electro-Fenton (EF), photoelectroFenton (PEF) and solar photoelectro-Fenton (SPEF)], because they show great ability to mineralize drugs and their metabolites from wastewater. Simpler electrochemical separation technologies, such as electrodialysis and electrocoagulation, already available at industrial scale, transfer only the pollutants out of the contaminated water. Fig. 1 presents the main analytical methods used to check the viability of electrochemical destruction technologies, such as the

EAOPs for water remediation. First, their performance may be assessed in terms of: 1 organics mineralization from the decay in total organic carbon (TOC) of the solution; 2 organics oxidation from the abatement of chemical oxygen demand (COD); and, 3 decay kinetics of initial pollutant, determined from reversedphase high-performance liquid chromatography (RP-HPLC). Note that GC can be used instead of HPLC whenever the nature of the pharmaceutical and/or its matrix allows simpler analysis. These results should comply with economic criteria, which are intimately related to the electrolysis time factor, since the electrical costs may determine the viability, so they are usually accompanied by estimated figures of merit [e.g. mineralization current efficiency (MCE) and energy consumption]. In this regard, the recent use of photovoltaic systems as a source of power for the electrochemical reactors opens the door to cut energy costs [8]. A second, crucial task concerns reactivity assessment, which is aimed at elucidating the reaction sequence undergone by the drugs, involving the detection of cyclic/aromatic products and shortchain aliphatic carboxylic acids using GC-mass spectrometry (MS) and LC-MS, as well as inorganic ions released by ion chromatography (IC). These analyses are related to technological criteria because knowledge of the gradual transformation of initial pollutants can optimize the integration of novel electrochemical methods on current WWTFs through their smart coupling with pre-existing technologies. Thus, the use of GC-MS and LC-MS becomes mandatory for this purpose. A third kind of study to evaluate the viability of an electrochemical treatment correctly is related to environmental criteria, and may be performed on the basis of the evolution of solution toxicity with time, which usually arises from the standard Microtox test or determination of biological oxygen demand after 5-day incubation (BOD5). This is particularly useful for effluents with chloride ions and real wastewater with a complex matrix, which can give rise to toxic residues. Some studies have also evaluated the time course of cytotoxicity and mutagenicity, as in the case of antineoplastics contained in clinical wastewater [9]. This article looks into the reactivity of single drugs in synthetic and real solutions, and mixtures of pharmaceuticals in effluents upon treatment by electrochemical methods to show the large potential of GC-MS and LC-MS to identify the resulting products. We also examine complementary information from HPLC and IC analyses. First, we present an overview of the fundamentals and the performance of the most important electrochemical methods used to degrade drugs to improve understanding of the recent expansion of this research field. 2. Electrochemical technologies for removing pharmaceuticals from water streams

Fig. 1. Analytical techniques to test the viability of new devices aimed at applying electrochemical degradation technologies. Performance assessment: TOC, Total organic carbon; COD, Chemical oxygen demand; HPLC, High-performance liquid chromatography. Reactivity assessment: IC, Ion chromatography; GC-MS, Gas chromatographymass spectrometry; LC-MS, Liquid chromatography-mass spectrometry. Toxicity assessment: Microtox method, based on determining the bioluminescence inhibition of the marine bacteria Vibrio fischeri; BOD5, Biological oxygen demand after 5-day incubation.

Over the past decade, EAOPs, such as EO, EF, PEF and SPEF, have shown high effectiveness for pharmaceuticals mineralization, as discussed below. We also comment on other electrochemical methods [e.g. photoelectrocatalysis (PEC), and emerging microbial fuel cells (MFCs)], along with coupled methods. Note that the interest in classical electrochemical treatment, such as electrocoagulation, is significantly lower due to its poor oxidation ability, which leads to the accumulation of most removed organics in the final sludge [1,10]. 2.1. Electrochemical oxidation EO, also called anodic oxidation when non-chloride solutions are treated, is the most widespread EAOP for water remediation. It

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

involves the oxidation of a contaminated solution in a divided or undivided electrochemical cell usually operating at constant current density (j). Organics can then be directly oxidized at anode M, but, when operating at high j values, they are pre-eminently destroyed by adsorbed M(•OH) formed from water oxidation [1,3,7]:

M + H2O → M ( i OH) + H+ + e−

3

a

1.0 0.8

(1)

0.6

M(•OH)

2 Cl− → Cl2 (aq ) + 2 e−

(2)

Cl2 (aq ) + H2O  HClO + Cl− + H+ (acid medium )

(3)

Cl2 (aq ) + 2 OH−  ClO− + Cl− + H2O (alkaline medium )

(4)

HClO  ClO− + H+

pK a = 7.55

(5)

In addition, toxic chlorine oxyanions (e.g. ClO2−, ClO3− and ClO4−) may be produced, although their content can be minimized by using low j values [7]. In this method, the so-called EO with active chlorine, organics are attacked by different species depending on solution pH, the predominant species being Cl2 until pH is near 3, HClO in the pH range 3–8 and ClO− at pH > 8 [1]. The mediated oxidation

0.4

TOC / TOC0

0.2 0.0

b

1.0 0.8 0.6 0.4 0.2 0.0

c

100 80

% MCE

depends on the nature of the The oxidation ability of anode along with operational variables (e.g. pH, substrate concentration, temperature and stirring or flow rate). According to their nature, the anodes are classified as active and non-active. An anode is active if M(•OH) is preferentially oxidized to an “active oxygen species” MO that only allows the electrochemical conversion of organics into short-chain carboxylic acids. Several pharmaceutical solutions have been degraded by EO using active anodes {e.g. Pt [11] and Ti/RuO2-IrO2 [12,13]}. In contrast, a non-active anode favors the formation of physisorbed M(•OH) that leads to the electrochemical mineralization of organics. By using this approach, the treatment of synthetic and real drug solutions by EO with SnO2 doped with Sb [13,14] and boron-doped diamond (BDD) [11,15–17] has been reported. BDD thin-film electrodes deposited onto p-Si, Nb or Ti show the highest overpotential for O2 evolution and the low adsorption enthalpy of BDD-•OH allows the production of larger amounts of reactive BDD(•OH) compared to other anodes, thereby yielding the most effective mineralization of organic pollutants. Other weaker reactive oxygen species (ROS) (e.g. O2•−, H2O2, HO2• and O3) can also be generated from water discharge at all anodes. In sulfate medium, the oxidizing peroxodisulfate ion can also be formed [1]. As an example, Figs. 2a and 2b depict the comparative decay of normalized TOC of 245 mg L−1 of the antibiotic chloramphenicol (2,2-dichloro-N-[1,3-dihydroxy-1-(4-nitrophenyl) propan-2-yl]acetamide) in 0.05 M Na2SO4 using 100 mL stirred Pt/air-diffusion and BDD/air-diffusion tank reactors with 3 cm2 electrode area [11]. While TOC remained almost unchanged when a Pt anode was used at 33.3 mA cm−2, 78% mineralization was achieved using a BDD anode for 360 min, in agreement with the much greater oxidation ability of BDD(•OH) compared to Pt(•OH). For EO with BDD, Fig. 2c highlights an MCE of about 16% during all the electrolysis, thus suggesting a constant mineralization rate of all oxidation products. In contrast, Fig. 3 reveals a slow, but gradual, destruction of chloramphenicol when a Pt anode was used, being reduced by 52% at 360 min, whereas the most potent BDD anode yielded its almost complete removal at the same time. The scarce TOC abatement using a Pt anode (Fig. 2a) can then be ascribed to the very low reactivity of Pt(•OH) to destroy the degradation products. A very different behavior is usually found when the treated effluent contains chloride ions, as in the case of real wastewater [1,7,12,13,15,17], since active chlorine species (e.g. Cl2, HClO and/ or ClO−) are generated from Cl− oxidation via reactions (2)–(5) as follows:

60 40 20 0 0

60

120

180

240

300

360

420

Electrolysis time / min Fig. 2. Normalized total organic carbon (TOC) abatement with electrolysis time for the ( ) electrochemical oxidation, ( ) electro-Fenton, ( ) photoelectroFenton with a 6 W UVA light (λmax = 360 nm) and ( ) solar photoelectro−1 Fenton treatments of 100 mL of 245 mg L of the antibiotic chloramphenicol in 0.05 M Na2SO4 at pH 3.0 using (a) Pt/air-diffusion and (b) boron-doped diamond (BDD)/ air-diffusion tank reactor at 33.3 mA cm−2 and 35°C. In the three latter methods, the solution contained 0.5 mM Fe2+ for Fenton’s reaction. In plot (c), the mineralization current efficiency for the above trials is shown [11].

with active chlorine becomes faster in acidic media than it is in alkaline media because of the higher standard potential of Cl 2 (E° = 1.36 V/SHE) and HClO (E° = 1.49 V/SHE) compared to ClO− (E° = 0.89 V/SHE). The major drawback of this procedure is that it is liable to generate and to accumulate toxic chloroderivatives, trihalomethanes and chloramines, which enhance the toxicity of the effluent. Moreover, the additional oxidative action of active chlorine does not usually entail higher COD or TOC removal rates

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

4

[Chloramphenicol] / mg L-1

250 200 150 100 50 0

0

60

120

180

240

300

360

420

Electrolysis time / min Fig. 3. Chloramphenicol concentration decay versus electrolysis time for the ( ) electrochemical oxidation and ( ) solar photoelectro-Fenton degradations under the conditions of Fig. 2 using a ( ) Pt/air-diffusion or ( ) BDD/air-diffusion tank reactor [11].

compared to those determined for EO in sulfate medium, although at least it accelerates the decay of the initial pharmaceutical and its cyclic/aromatic products, as found for β-blocker metoprolol [13,15]. 2.2. Electro-Fenton and photoelectro-Fenton Indirect electro-oxidation methods based on Fenton’s reaction chemistry (e.g. EF, PEF and SPEF) are based on the continuous generation of H2O2 from reduction of injected O2 at a carbonaceous airdiffusion [11] or carbon-felt cathode [18], from reaction (6). The oxidation power of H2O2 is further enhanced in the presence of a small quantity of added Fe2+ ion to yield Fe3+ ion and homogeneous •OH in the bulk from Fenton’s reaction (7) at optimum pH 2.8 [1].

O2 + 2 H+ + 2 e− → H2O2

(6)

Fe2+ + H2O2 + H+ → Fe3+ + H2O + iOH

(7)

These techniques are operative for the treatment of acidic aqueous pharmaceutical effluents to avoid the precipitation of catalytic iron ions as hydroxides [1,7,11,18–22]. Since undivided cells are often used, organics are simultaneously destroyed by M(•OH) formed from reaction (1) and •OH generated from Fenton’s reaction (7). However, hydroxyl radicals are also consumed by parasitic reactions that reduce the oxidizing power and the efficiency of the systems [7]. In PEF, the solution treated under EF conditions is illuminated with artificial UV light, usually in the range 320–400 nm with λmax = 360 nm. The results of this radiation are beneficial for the degradation process because it causes: 1 the reductive photolysis of [Fe(OH)]2+, the pre-eminent form of Fe3+ ion at pH near 3, thus regenerating Fe2+ ion and producing additional •OH by photoreduction reaction (8); and, 2 the photolysis of some products or their Fe(III) complexes promoting Fe2+ regeneration, as in the case of photodecarboxylation of Fe(III)-carboxylate species via reaction (9) [1,20–22].

[Fe (OH)]2+ + hν → Fe2+ + iOH

(8)

Fe (OOCR ) + hν → Fe2+ + CO2 + R i

(9)

2+

The major drawback of PEF in practice is the high energy requirements from UV lamps. To solve this, our group recently proposed the alternative use of sunlight as a renewable, inexpensive energy source in the SPEF process [8,11,19]. Figs. 2a and 2b show the increasing abatement of normalized TOC for the 245 mg L−1 chloramphenicol solution in sulfate medium at pH 3, when the oxidation power of the process rises in the sequence EO < EF < PEF (with a 6 W UVA lamp) < SPEF for both stirred Pt/air-diffusion and BDD/air-diffusion tank reactors, respectively [11]. The higher oxidation ability of EF compared to EO was explained by the quicker destruction of organics under the action of additional •OH generated from Fenton’s reaction (7), which was further enhanced in PEF by the photolysis of final Fe(III)-carboxylate complexes. The superiority of SPEF related to the much greater UV intensity of sunlight. In all EAOPs, a better performance was found with BDD anode. Thus, SPEF allowed the quickest mineralization, attaining 100% TOC removal in 180 min and 120 min using Pt and BDD anodes, respectively. This is also reflected in Fig. 2c, where the highest MCE values correspond to the latter two processes and maximal values of about 80% and 100% at 60 min were reached, respectively. The dramatic decay in efficiency at longer time was ascribed to the drop in organic matter content and their more recalcitrant nature. However, similar and very fast chloroamphenicol abatement was found for EF, PEF and SPEF, with total disappearance in ~25 min, as can be seen in Fig. 3. This indicates that the initial drug and its aromatic intermediates are much more rapidly destroyed by • OH in the bulk than by heterogeneous Pt( • OH) or BDD(•OH), whereas the generation of photoinduced •OH from reaction (8) is rather insignificant. From the above results, the viability of EAOPs, especially Fentonbased EAOPs, to treat pharmaceutical residues and their reaction products, is demonstrated, although, in order to foster their integration with other technologies applied in WWTFs, the nature and the toxicity of such intermediates must be known. 2.3. Other electrochemical methods Photoelectrocatalysis involves the application of a constant bias anodic potential (Eanod) or a constant current to a semiconductorbased thin-film anode subjected to UV illumination. This produces positively-charged holes and photoexcited electrons that are continuously extracted from the anode and injected into the cathode [1]. Organics are then primordially oxidized by 1 photogenerated holes; or, 2 •OH formed from water oxidation caused by the holes. Only a few papers have reported the treatment of pharmaceuticals solutions by PEC due to its low oxidation power, being particularly useful for organic micropollutants [23–26]. For example, Liu et al. [23] treated 25 mL of synthetic wastewater with 50 mg L−1 of the antibiotic tetracycline in 0.02 M Na2SO4 at pH 4.5 using a TiO2nanotube-array photoanode at Eanod = 0.5 V/SCE illuminated with a 4 W UV lamp of λmax = 254 nm, with 81% of the drug being removed in 180 min. In contrast, Daghir et al. [25] reported the total abatement of chlortetracycline with 92% TOC removal when 600 mL of 100 μg L−1 of this drug in 0.07 M Na2SO4 were treated for 180 min by PEC at 600 mA using a 24 cm2 N-doped Ti/TiO2 photoanode exposed to sunlight. This suggests the potential use of sunlight as a radiation source in PEC, although more research is required to demonstrate its viability. Recently, several authors reported the possible use of microbial fuel cells (MFCs) equipped with anodic biofilms to degrade pharmaceutical effluents. Liu et al. [27] proposed an MFC configuration with carbon electrodes, fed with a steroidal drug-production wastewater and with electricity co-generation. Removal of 82% COD,

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

62% N and 26% sulfate, with 22.3 W m−3 power density and 30% coulombic efficiency was obtained as maximal. Harnish et al. [28] described the formation of biofilms on a graphite anode by applying an Eanod = 0.2 V/Ag/AgCl, with the ability to remove almost completely 6 μg L−1 of sulfonamides (e.g. sulfamethoxazole, sulfathiazole, sulfadimidine and sulfadiazine) and their corresponding N4-acetyl metabolites. As mentioned above, coupling of electrochemical processes with existing technologies should be the right way towards their implementation at industrial scale. Urtiaga et al. [29] have devised a prepilot system, coupling ultrafiltration, reverse osmosis and EO with a BDD anode, to assess the removal of 77 pharmaceuticals detected as micropollutants in a secondary WWTF effluent. While ultrafiltration only reduced the initial organic load by about 20%, reverse osmosis was more efficient giving 50–70% removal of pharmaceuticals. Subsequent application of EO allowed their almost total removal, which became more effective by increasing j. These results highlight the large potential of membrane filtration coupled to electrochemical methods for the effective remediation of real pharmaceutical wastewater. 3. Elucidation of reaction products by mass spectrometry Several works have reported the reactivity of single pharmaceuticals in synthetic aqueous solutions upon treatment by EAOPs. Reaction products are usually determined by GC-MS analysis of electrolyzed effluents [21,22,30–41], although, recently, authors tended to use LC-MS as an alternative analytical technique [11,30,42–47] because LC-MS may allow detection of a larger number of products, even ionic ones. Unfortunately, literature on the application of LC-MS to real pharmaceutical effluents is rather scarce [12,13]. The main results found from both techniques are discussed for various EAOPs tested, and the reactivity of each drug is also thoroughly elucidated by considering the products simultaneously identified by HPLC and IC. 3.1. GC-MS The lack of volatility and/or the poor thermal stability of some pharmaceuticals and/or their reaction products could limit the application of GC-MS. Organic solvents of low boiling point (e.g. CH2Cl2) are usually employed to extract the non-ionic organic compounds formed during electrolysis. This represents another important drawback of such methodology because only a partial extraction of organics is achieved, whereas some non-ionic and most ionic compounds remain in the aqueous phase. Solid-phase extraction (SPE) can be utilized when very low concentrations of products are expected, although SPE is not widespread in the studies we reviewed for this article. When the organic components contain –OH groups, they can be derivatized by their reaction with N,O-bis-(trimethylsilyl)acetamide (silylation), ethanol (ethylation) or acetyl chloride (acetylation) to identify a greater number of intermediates, especially those showing low molar masses. Polar and non-polar columns are used to separate the organics in the gas chromatograph and they are identified from their mass spectra upon structural determination with support from libraries (e.g. NIST 14 MS) [48]. Zhao et al. [30] reported the formation of chloroderivatives for diclofenac (sodium [2-[(2,6-dichlorophenyl)amino]phenyl]acetate) mineralization using EO with active chlorine and a BDD anode. GC-MS analysis revealed the generation of benzoic acid, 2,6dichlorobenzenamine and 2,5-dihydroxybenzyl alcohol, which were transformed into carboxylic acids (e.g. hexanoic, butanedioic, malonic, 2-hydroxylpropanoic, 2-hydroxylacetic and oxalic). These products were identified upon silylation, and GC-MS was carried out on an Agilent 6890GC/5973MSD with a DB-5 MS column.

5

Since hydroxyl radicals are the main oxidizing agents in all the EAOPs when they are carried out in chloride-free solutions, analogous oxidation products are always expected for each pharmaceutical, although their accumulation-degradation profiles will depend on the electrodes used, the reactor design, the operation variables and/or the action of UV light or sunlight in photo-assisted processes. One of the simplest drugs degraded by these methods has been the antiinflammatory ibuprofen (2-(4-isobutylphenyl)propionic acid), with no heteroatoms in its structure [34]. A saturated solution with 41 mg L−1 of this drug in 0.050 M Na2SO4 at pH 3.0 and 0.50 mM Fe2+ was treated by EF, PEF (with a 6 W UVA light) and SPEF using 100 mL stirred Pt/O2-diffusion and BDD/O2-diffusion cells. Aromatic products were directly extracted with CH2Cl2 and detected by GC-MS using a Fisons GC 8060 with a polar HP Innowax 0.25 μm column and coupled to a Fisons Thermo Finnigan MD800 MS operating in electron impact (EI) mode at 70 eV and 300°C. Two degradation pathways were proposed. The first involved the hydroxylation of its propionic acid group, consecutively yielding 1-(1-hydroxyethyl)-4isobutylbenzene, 4-isobutylacetophenone and 4-isobutylphenol, whereas, in the second, both propionic acid and isobutyl substituents were oxidized to 4-ethylbenzaldehyde. Further reactivity of hydroxylated products depends on the steric hindrance and the existing functional groups that activate and deactivate the positions of the ring and its substituents. Thus, GC-MS analysis upon treatment of paracetamol (N-(4-hydroxyphenyl)acetamide) by EO, EF and PEF using a Pt/O2-diffusion cell revealed the formation of primary product hydroquinone and its transformation product p-benzoquinone [31]. In similar treatments of another N-containing drug, β-blocker metoprolol (1-(isopropylamino)-3-[4-(2-methoxyethyl)phenoxy]-2propanol), 4-(2-methoxyethyl)phenol was formed, followed by successive oxidation to methyl 4-hydroxyphenylacetate, 2-hydroxy-2(4-hydroxyphenyl)acetic acid and 4-hydroxybenzaldehyde [36]. For drugs with chlorine atoms, chlorinated aromatic products may be detected by GC-MS. For example, the antibiotic triclosan (2,4,4’-trichloro-2’-hydroxydiphenyl ether) treated by EF using a Pt or a BDD anode and a carbon-felt or O2-diffusion cathode yielded hydroxylated products (e.g. 2,4-dichlorophenol, 4-chlorocatechol and chlorohydroquinone) and chloro-p-benzoquinone [32]. In contrast, in some occasions, the •OH attacks the –Cl position and no chloroderivatives are found in solution. For example, biocide chloroxylenol (4-chloro-3,5-dimethylphenol) was transformed into 2,6-dimethylhydroquinone, 2,6-dimethyl-p-benzoquinone and 3,5dimethyl-2-hydroxy-p-benzoquinone by EO, EF and PEF in similar cells [33]. An interesting work by Guinea et al. [35] reported excellent complementary results from HPLC and GC-MS analyses of solutions of the veterinary fluoroquinolone antibiotic enrofloxacin electrolyzed by EO, EF, PEF and SPEF. Table 1 summarizes the name and the characteristics of the four primary aromatic products detected by RP-HPLC, corresponding to two hydroxylated derivatives of the initial drug and two compounds produced from the alternative attack of hydroxyl radicals on the piperazine side chain, yielding its partial cleavage or causing the appearance of an –NH2 group. Table 2 collects 17 short intermediates, such as polyols, ketones, carboxylic acids and N-derivatives formed upon degradation of the above aromatics and identified by different GC-MS conditions, and seven final carboxylic acids that were also detected by ion-exclusion HPLC. These results show the most typical reactivity of pharmaceuticals degraded by EAOPs, involving the initial attack of hydroxyl radicals to give hydroxylated and other oxidized products, which are subsequently transformed into short intermediates that evolve to final short-linear aliphatic acids (e.g. oxalic and oxamic) [1,7]. Consequently, RP-HPLC can often be used not only to identify, but also to quantify early degradation products, primarily when they are relatively large and polar, whereas GC-MS or ion-exclusion HPLC is more suitable for the detection of much less concentrated products

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

6

Table 1 Primary aromatic compounds detected in the RP-HPLCa chromatograms recorded during the EO, EF, PEF and SPEF degradation of the fluoroquinolone enrofloxacin using stirred tank reactors equipped with a Pt or BDD anode and an O2-diffusion cathode [35]. Aromatic compound

Molecular structure

Enrofloxacin (1-cyclopropyl-7-(4-ethyl-1-piperazinyl)-6-fluoro-1,4-dihydro-4-oxo-3-quinolinecarboxylic acid)

O

Retention time (min) 16.96

O F

HO N

N N

O

1-cyclopropyl-7-(4-ethyl-1-piperazinyl)-1,4-dihydro-6-hydroxy-4-oxo-3-quinolinecarboxylic acid

O

5.98 OH

HO N

N N

O

1-cyclopropyl-7-{[2-(ethylamino) ethyl]amino}-6-fluoro-1,4-dihydro-4-oxo-3-quinolinecarboxylic acid

O

9.66 F

HO N

7-amino-1-cyclopropyl-6-fluoro-1,4-dihydro-4-oxo-3-quinolinecarboxylic acid

O

NH

11.60

O F

HO N

O

1-cyclopropyl-7-(4-ethyl-1-piperazinyl)-6-fluoro-1,4-dihydro-8-hydroxy-4-oxo-3-quinolinecarboxylic acid

H N

NH2

O

15.18 F

HO

N

N OH

N

a Analysis performed using a Waters 600 LC equipped with a Thermo BDS Hypersil 5 μm C18 column and coupled with a Waters 996 photodiode-array detector. The mobile phase was a mixture of 13:87 (v/v) acetonitrile/2.45 g L−1 H3PO4 at pH 3.0 with triethylamine at 1.5 mL min−1.

accumulated upon several degradation steps or final carboxylic acids, respectively. 3.1.1. Use of HPLC, IC and toxicity analysis to clarify the reactivity of pharmaceuticals RP-HPLC analysis of electrolyzed solutions has been employed to analyze the kinetic behavior of pharmaceuticals and, sometimes, also that of their primary products. For example, in the study on ibuprofen [34], this characterization technique was performed under similar conditions to those given in Table 1 and revealed a pseudo-first-order decay of ibuprofen concentration in all cases. Under EO conditions using cells with a stainless-steel cathode, the drug was removed much more rapidly with a BDD anode than a Pt anode, as expected by the higher oxidation ability of BDD(•OH) compared to Pt(•OH). Ibuprofen removal was strongly accelerated in EF, being much more rapid for PEF and SPEF, regardless of the anode used, thus confirming the superior oxidation ability of •OH formed in the bulk from Fenton’s reaction (7). The same behavior was also observed in RP-HPLC for the evolution of the intermediate 4-isobutylacetophenone, which persisted in each EAOP while ibuprofen was still present. The superiority of homogeneous •OH over BDD(•OH) or Pt(•OH) to destroy aromatic drugs and their aromatic products, and the total removal of all these molecules practically at the same time, are general rules verified for all the EAOPs operating in sulfate medium. As mentioned above, the progressive oxidative cleavage of aromatic products of pharmaceuticals leads to final short-linear aliphatic carboxylic acids, which can be quantified by ion-exclusion HPLC using, for example, a liquid chromatograph fitted with an Aminex HPX 87H column and coupled with a photodiode-array detector set

at λ = 210 nm, under circulation of simple mobile phases (e.g. 4 mM H2SO4) [35]. Malic, maleic, fumaric and acetic acids are commonly detected, being further oxidized to oxalic and formic acids [1,7,31–36]. When the drug contains N atoms, oxamic acid coming from the degradation of N-derivatives (e.g. acetamide) may also be formed [22,31,37,41]. Oxalic, oxamic and formic acids are directly converted into CO2 [7]. The great persistence of the latter acids explains the long duration of EAOPs in removing solution TOC. Nevertheless, their removal rate is a function of the anode and the EAOP used. In EO, the final acids can be destroyed by BDD(•OH) but not by Pt(•OH). In contrast, the addition of iron ions in EF, PEF and SPEF leads to the formation of Fe(II)-carboxylate complexes using a carbon-felt cathode or Fe(III)-carboxylate complexes using a gasdiffusion cathode [32]. While the former species can be degraded by •OH in the bulk, the latter ones are more refractory. The quick photolysis of Fe(III)-carboxylate species, primarily of the major Fe(III)oxalate complexes, under UVA light or sunlight in PEF or SPEF, respectively, explains the better performance of these methods in decontamination of water with pharmaceuticals [22,41]. The addition of Cu2+ as co-catalyst in EF and PEF accelerates the destruction of carboxylic acids due to the parallel destruction of Cu(II)carboxylate species [31]. Regarding the presence of heteroatoms (e.g. Cl, S and F in the pharmaceuticals), inorganic ions (e.g. Cl−, SO42− and F−) are accumulated in the treated solutions [33,35,36,38], although the former ion is slowly oxidized to chlorine by reaction (2) using a BDD anode. In contrast, the N atom tends to be preferentially converted into NH4+ ion, along with a small proportion of NO3− ion, although their mass balance reveals the loss of N-volatile compounds (e.g. NOx and N2)

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

7

Table 2 Aliphatic carboxylic acids and other short intermediates identified by GC-MS and ion-exclusion HPLC during the treatments of enrofloxacin reported in Table 1 [35]. Compound

Molecular structure

Analytical technique

Retention time (min)

Molecular mass (g mol-1)

1,3,4,5,6-Pentahydroxyhexan-2-one

HO

GC-MSa

12.61

437

OH

HO

OH OH

O

OH

Ethylene glycol

GC-MSb,c

3.31

118b, 146c

GC-MSb

8.57

161

GC-MSb HPLC

7.58 11.8

174

GC-MSb

8.44

191

HPLC

8.4

-

GC-MSa HPLC

6.67 8.1

260

GC-MSa HPLC

6.67 15.1

260

GC-MSb HPLC

6.41 9.8

160

GC-MSb HPLC

5.54 6.5

146

GC-MSb HPLC

6.33 9.2

117

GC-MSb

5.82

131

GC-MSb

3.92

89

GC-MSa,d

4.52

117a

GC-MSd

3.87

59

GC-MSb

6.58

88

GC-MSd

6.20

73

GC-MSd

19.09

73

HO

6-Hydroxy-5-oxo-hexanoic acid

HO OH O

O

Succinic acid

O HO OH O O

Malic acid HO

OH O

OH

Tartaric acid

O

OH HO

OH O

OH

OH

Maleic acid

O

O

OH

Fumaric acid

O OH HO O

Malonic acid

O

O

HO

Oxalic acid

OH

HO

O

O

OH

O

Oxamic acid

O

H 2N

Acetylcarbamic acid

OH

O

O

HO

N H

Carbamic acid

O

HO

NH2 O

Acetic acid HO

Acetamide H2N O

Ethane diamide

O NH2 H2N O

Diethyl amine

Ethyl formamide

N H

O

N H

GC-MS analyses were performed after electrolyzing 100 mL of 158 mg L−1 enrofloxacin solutions in 0.05 M Na2SO4 of pH 3.0 at 33 mA cm−2 for: a180 min by electrochemical oxidation with Pt followed by silylation, b180 min by electro-Fenton with Pt and 0.5 mM Fe2+ followed by ethylation, c60 min by electrochemical oxidation with Pt followed by acetylation and d180 min by electrochemical oxidation with Pt without any derivatization. The molecular mass corresponds to either the trimethylsilyl, ethyl or acetyl derivative or the raw intermediate, respectively. The analyses were performed with a Hewlett-Packard 5890 Series II GC, using either a polar HP-Innowax 0.25 μm column or a nonpolar HP-5MS one with 5% phenyl methyl siloxane, coupled to a Hewlett-Packard 5989A MS operating in EI mode at 70 eV.

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

8

[22,31]. All these ions can be accurately quantified by IC, using specific anion (e.g. Shim-Pack IC-A1S) or cation (e.g. Shodex IC KY421) columns and a conductivity detector [36]. Lately, special concerns have arisen from the presence of antibiotics in the environment because of their bioaccumulative properties that cause unpredictable but irreversible changes, such as the growth of antimicrobial-resistant microorganisms [37]. The latest World Health Organization report on global surveillance of antimicrobial resistance in 2014 stated that it threatens the effective prevention and treatment of an ever-increasing range of infections [49]. However, there is a lack of information on the correlation between degradation products and their potential contribution to antimicrobial resistance. This aspect was considered in a recent article of Dirany et al. [37], who correlated the time course of both the antibiotic sulfachloropyridazine ((4-amino-N-(6-chloro-3-pyridazinyl) benzenesulfonamide) and its products generated in EF with Pt/ carbon-felt and BDD/carbon-felt cells, with toxicity profiles determined by Microtox. Up to 10 aromatic intermediates were detected by GC-MS and it was found that their production at the beginning of electrolysis strongly enhanced the toxicity of the solution. Comparative EF treatments of the two main products, amino6-chloropyridazine and p-benzoquinone, confirmed their higher toxicity compared to the initial antibiotic. The formation of more toxic hydroxylated products from sulfonamides treated by EAOPs is usual. Thus, catechol, resorcinol, hydroquinone and p-benzoquinone have been identified by GC-MS during the degradation of sulfanilamide (4-aminobenzenesulfonamide) [22] and sulfamethazine (4amino-N-(4,6-dimethyl-2-pyrimidinyl)benzene sulfonamide) [38] by EF, PEF and SPEF with Pt/air-diffusion and BDD/air-diffusion tank reactors. For other sulfonamides, Fabianska et al. [40] reported the generation of urine derivatives, sulfanilic acid and hydroquinone by EO with BDD. 3.2. LC-MS LC-MS has the advantage of detecting more products from pharmaceuticals because they can be directly separated from the treated aqueous solution. The first work dealing with the use of LC-MS for the determination of degradation products of pharmaceuticals upon treatment by EAOPs was reported by Zhao et al. [30], who degraded a 30 mg L−1 diclofenac solution in 0.1 M NaCl of pH 6.7 by EO with active chlorine. Chlorinated products (e.g. dichlorodiclofenac) were detected by means of LC/electrospray-time-of-flight massMS (LC/ESI-TOF-MS). Other authors have also explored the potential of the combined use of LC-MS and GC-MS (e.g. to identify the products of different sulfonamides) [37,40]. An interesting work of Radjenovic et al. [12] described the disappearance of 28 common pharmaceuticals and pesticides spiked in trace concentrations (7.8–37.4 μg L−1) in a 10 L reverse osmosis concentrate wastewater of pH 7.5 treated by EO using a divided flow cell equipped with a 24 cm2 Ti/Ru0.7Ir0.3O2 anode and a 24 cm2 stainless-steel cathode. Total removal of all pollutants was achieved after consuming 437 A h m-3 operating in batch, with a loss of 25% of the initial TOC of the effluent. Destruction of organics, along with natural organic matter (NOM), was ascribed to mediated oxidation by electrogenerated oxidants [e.g. ROS (•OH, HO2• and H2O2) and reactive halogenated species (RHS) including active chlorine (HClO/ ClO−)] and other radicals (e.g. Br•, Br2•−, Cl•, and Cl2•−). Based on these results, the same authors studied the EO degradation of 50 μM metoprolol (MTPL) in a reverse osmosis concentrate using the previous electrolytic system with a Ti/Ru0.7Ir0.3O2 or Ti/SnO2-Sb anode [13]. Using Oasis HLB cartridges to enhance the extraction of organics by SPE, up to 22 aromatic products (P) were detected by LCMS with both anodes. Fig. 4 shows the complex pathway proposed involving the cleavage of lateral groups (P225, P237, P253) and hydroxylation (P279, P281, P283) with ROS, along with the formation

of monochloro- (P271, P301, P313, P315, P317), dichloro- (P293, P 335, P347, P349), trichloro- (P381, P383), monobromo- (P345, P357, P359) and monochloro-monobromo- (P379) derivatives under the action of RHS. However, a detrimental effect of the presence of these halogenated products was the significant increase of toxicity. For a metoprolol solution without halide ions, oxidation by ROS during EO with BDD anode allowed the identification of similar intermediates [45]. Intermediates formed during the EO treatment of other pharmaceuticals with a BDD anode have also been detected by LC-MS. Martín de Vidales et al. [42] found partial mineralization of 500 mL of 0.1–100 mg L−1 sulfamethoxazole (4-amino-N-(5-methylisoxazol3-yl)-benzenesulfonamide) in 5 g L−1 Na2SO4 using a bench-scale flow plant. Significant amounts of nine intermediates were identified by using the SPE cartridges mentioned above, arising from various routes [e.g. monohydroxylation on the benzenic ring (m/z = 270.0543), dihydroxylation on the isoxazol moiety (m/z = 286.0503), hydrolysis to yield 5-methylisoxazol-3-amine (m/z = 99.0553), 4H-1,2,4benzothiadiazine-6-nitro-1,1-dioxide (m/z = 225.9928), 4-aminoN-(iminomethylene)-benzenesulfonamide (m/z = 225.9928) and 4-aminobenzenesulfonamide (m/z = 173.1377), and further oxidation of these intermediates to yield 4-aminobenzenesulfonic acid (m/z = 198.0331), its nitro derivative 4-nitrobenzenesulfonic acid (m/z = 201.9818) and p-benzoquinone (m/z = 109.0284)]. Our group has also reported the partial mineralization of the drug omeprazole {5-methoxy-2-{[(4-methoxy-3,5-dimethyl-2-pyridinyl) methyl]sulfinyl}-1H-benzimidazole} in a phosphate buffer of pH 7.0 using EO with a BDD anode [45]. Eight aromatic products were identified by LC-MS using a Shimadzu SIL-20AC LC equipped with a Teknokroma Mediterranean Sea C-18 column and coupled to a Shimadzu LCMS-2020 MS. These products related to the cleavage of lateral groups or the release of benzene, benzimidazole or pyridine moieties. In our above study on chloramphenicol [11], nine aromatic products, 13 hydroxylated derivatives and dichloroacetic acid were found by LC-MS in all the EAOPs tested. Moreover, oxalic, oxamic and formic acids were detected as final carboxylic acids and the quick photolysis of their Fe(III) complexes accounted for the superiority of SPEF with Pt or BDD (see Fig. 2). The initial Cl and N atoms were converted into Cl− and NO3− ions, respectively. The detection of intermediates by LC-MS in Fenton-based processes has also been performed by Oturan’s group. Up to seven and eight aromatic products were identified during the treatment of the antibiotics tetracycline [43] and ciprofloxacin [47], respectively, with a Ti/RuO2-IrO2 or Pt anode and a carbon-felt cathode. Furthermore, the toxicity of the tetracycline solution measured from the immobilization of the crustacean Daphnia magna was shown to increase largely at the beginning of EF, indicating once again that great attention has to be paid to the time course of toxicity during the electrochemical treatment of pharmaceutical wastewater. 4. Conclusion and outlook Two electrochemical technologies, namely SPEF and EO, exhibit the greatest performance for the treatment of water containing pharmaceuticals, although new directions in the field still need to address a couple of key aspects. Regarding SPEF, perspectives at mid-term focus on solar power for cost reduction. Autonomous pilot plants were devised in recent years for other electrochemical technologies (e.g. EO, EC and electrodialysis), but a solar-powered SPEF system was recently tested in our laboratory. Promising results obtained in all cases show the trend toward self-sustainability of electrochemical water treatments. As for EO, although most recent studies focus on its scale-up, some research is still performed at the laboratory scale. Despite the great impact that the appearance of the BDD anode had in the field due to its outstanding characteristics and performance, extremely

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

RHS P237

P271

ROS

ROS

9

RHS

RHS MTPL

P253

P345

RHS

RHS P293

P379

P301

P225

P335

P133 P279 (R: =O)

P281 (R: -OH) P281-II (R: =O)

P283-I

P283-II (R: -OH)

P315-II (R: =O) P317-II (R: -OH)

P357 (R: =O)

P359 (R: -OH)

RHS

RHS

P313 (R: =O)

P347 (R: =O)

P381 (R: =O)

P315 (R: -OH)

P349 (R: -OH)

P383 (R: -OH)

Fig. 4. Proposed reaction scheme for the electrochemical oxidation of beta-blocker metoprolol (MTPL) contained in a reverse osmosis concentrate using a Ti anode coated with a Ru0.7Ir0.3O2 or an SnO2-Sb metal-oxide layer. The products (P) were characterized by LC-MS and by MS2 and MS3 experiments by applying collision-induced dissociation (CID). ROS, Reactive oxygen species; RHS, Reactive halogen species. Analyses were performed using a Shimadzu Prominence ultra-fast LC equipped with an Alltima C18 column and coupled with a 4000 QTRAP quadrupole-linear ion trap MS (QqLIT-MS) composed of a Turbo Ion Spray Source [13].

high price and large energy consumption limit their possible applications at the industrial scale. The search for sufficiently potent, less expensive new anode materials is therefore a thriving research area that opens the door to enhanced, stable, resistant coatings (e.g. dimensionally-stable anodes). As clearly stated in this review, gaining knowledge on the reactivity of pharmaceuticals through elucidation of degradation products is essential for the correct exploitation of these technologies in practice. Apart from the use of GC-MS and LC-MS as described in this work, opportunities can also arise from direct detection of intermediates and/or side products through the use of differential electrochemical mass spectrometry (DEMS), although it has not yet been utilized for pharmaceuticals degradation. This could be particularly interesting for addressing a complete mass balance upon treatment, which could then include both volatile organic hydrocarbons and nitrogenated (e.g. NOx, N2) or chlorinated (e.g. CHCl3) species. DEMS is based on the use of miniaturized electrochemical cells composed of a porous anode, usually of Pt deposited on a PTFE membrane, in contact with a vacuum to extract the volatile products, even at the nM scale, to be analyzed by MS. More information on this technique was thoroughly reviewed by Baltruschat [50]. Finally, further work on the exact role of ROS and RHS on reaction pathways should consider in greater detail the effect of real pharmaceutical water matrices, including interferences from NOM (e.g. humic and fulvic acids) and radical scavengers (e.g. sulfate and carbonate ions and ironchelating species), which can eventually lower performance of the water treatment technology to be implemented.

Acknowledgements The authors are grateful to MINECO (Ministerio de Economía y Competividad, Spain) for financial support under project CTQ201348897-C2-1-R, co-financed with FEDER funds.

References [1] I. Sirés, E. Brillas, Remediation of water pollution caused by pharmaceutical residues based on electrochemical separation and degradation technologies: a review, Environ. Int. 40 (2012) 212–229. [2] I. Sirés, P.L. Cabot, F. Centellas, J.A. Garrido, R.M. Rodríguez, C. Arias, et al., Electrochemical degradation of clofibric acid in water by anodic oxidation. Comparative study with platinum and boron-doped diamond electrodes, Electrochim. Acta 52 (2006) 75–85. [3] L. Feng, E.D. van Hullebusch, M.A. Rodrigo, G. Esposito, M.A. Oturan, Removal of residual anti-inflammatory and analgesic pharmaceuticals from aqueous systems by electrochemical advanced oxidation processes. A review, Chem. Eng. J. 228 (2013) 944–964. [4] J. Rivera-Utrilla, M. Sánchez-Polo, M.A. Ferro-García, G. Prados-Joya, R. Ocampo-Pérez, Pharmaceuticals as emerging contaminants and their removal from water. A review, Chemosphere 93 (2013) 1268–1287. [5] J.O. Tijani, O.O. Fatoba, L.F. Petrik, A review of pharmaceuticals and endocrinedisrupting compounds: sources, effects, removal, and detections, Water Air Soil Pollut. 224 (2013) 1770–1798. [6] J. Zhang, V.W.C. Chang, A. Giannis, J.-Y. Wang, Removal of cytostatic drugs from aquatic environment: a review, Sci. Total Environ. 445–446 (2013) 281– 298. [7] I. Sirés, E. Brillas, M.A. Oturan, M.A. Rodrigo, M. Panizza, Electrochemical advanced oxidation processes: today and tomorrow. A review, Environ. Sci. Pollut. Res. 21 (2014) 8336–8367.

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013

ARTICLE IN PRESS 10

E. Brillas, I. Sirés/Trends in Analytical Chemistry ■■ (2015) ■■–■■

[8] S. Garcia-Segura, E. Brillas, Advances in solar photoelectro-Fenton: decolorization and mineralization of the Direct Yellow 4 diazo dye using an autonomous solar pre-pilot plant, Electrochim. Acta 140 (2014) 384–395. [9] J. Hirose, F. Kondo, T. Nakano, T. Kobayashi, N. Hiro, Y. Ando, et al., Inactivation of antineoplastics in clinical wastewater by electrolysis, Chemosphere 60 (2005) 1018–1024. [10] S. Farhadi, B. Aminzadeh, A. Torabian, V. Khatibikamal, M. Alizadeh Fard, Comparison of COD removal from pharmaceutical wastewater by electrocoagulation, photoelectrocoagulation, peroxi-electrocoagulation and peroxi-photoelectrocoagulation processes, J. Hazard. Mater. 219–220 (2012) 35–42. [11] S. Garcia-Segura, E.B. Cavalcanti, E. Brillas, Mineralization of the antibiotic chloramphenicol by solar photoelectro-Fenton. From stirred tank reactor to solar pre-pilot plant, Appl. Catal. B Environ. 144 (2014) 588–598. [12] J. Radjenovic, A. Bagastyo, R.A. Rozendal, Y. Mu, J. Keller, K. Rabaey, Electrochemical oxidation of trace organic contaminants in reverse osmosis concentrate using RuO2/IrO2-coated titanium anodes, Water Res. 45 (2011) 1579–1586. [13] J. Radjenovic, B.I. Escher, K. Rabaey, Electrochemical degradation of the β-blocker metoprolol by Ti/Ru0.7Ir0.3O2 and Ti/SnO2-Sb electrodes, Water Res. 45 (2011) 3205–3214. [14] D. Santos, M.J. Pacheco, A. Gomes, A. Lopes, L. Ciríaco, Preparation of Ti/Pt/SnO2– Sb 2 O 4 electrodes for anodic oxidation of pharmaceutical drugs, J. Appl. Electrochem. 43 (2013) 407–416. [15] M.J. Martín de Vidales, C. Sáez, P. Cañizares, M.A. Rodrigo, Metoprolol abatement from wastewaters by electrochemical oxidation with boron doped diamond anodes, J. Chem. Technol. Biotechnol. 87 (2012) 225–231. [16] J.R. Domínguez, T. González, P. Palo, J. Sánchez-Martín, M.A. Rodrigo, C. Sáez, Electrochemical degradation of a real pharmaceutical effluent, Water Air Soil Pollut. 223 (2012) 2685–2694. [17] P. Palo, J.R. Domínguez, T. González, J. Sánchez-Martin, E.M. Cuerda-Correa, Feasibility of electrochemical degradation of pharmaceutical pollutants in different aqueous matrices: optimization through design of experiments, J. Environ. Sci. Health A Tox. Hazard. Subst. Environ. Eng. 49 (2014) 843–850. [18] I. Sirés, N. Oturan, M.A. Oturan, Electrochemical degradation of β-blockers. Studies on single and multicomponent synthetic aqueous solutions, Water Res. 44 (2010) 3109–3120. [19] E. Isarain-Chávez, R.M. Rodríguez, P.L. Cabot, F. Centellas, C. Arias, J.A. Garrido, et al., Degradation of pharmaceutical beta-blockers by electrochemical advanced oxidation processes using a flow plant with a solar compound parabolic collector, Water Res. 45 (2011) 4119–4130. [20] S. García-Segura, J.A. Garrido, R.M. Rodríguez, P.L. Cabot, F. Centellas, C. Arias, et al., Mineralization of flumequine in acidic medium by electro-Fenton and photoelectro-Fenton processes, Water Res. 46 (2012) 2067–2076. [21] M.D.G. de Luna, M.L. Veciana, C.-C. Suc, M.-C. Luc, Acetaminophen degradation by electro-Fenton and photoelectro-Fenton using a double cathode electrochemical cell, J. Hazard. Mater. 217–218 (2012) 200–207. [22] A. El-Ghenymy, P.L. Cabot, F. Centellas, J.A. Garrido, R.M. Rodríguez, C. Arias, et al., Mineralization of sulfanilamide by electro-Fenton and solar photoelectroFenton in a pre-pilot plant with a Pt/air-diffusion cell, Chemosphere 91 (2013) 1324–1331. [23] Y. Liu, X. Gan, B. Zhou, B. Xiong, J. Li, C. Dong, et al., Photoelectrocatalytic degradation of tetracycline by highly effective TiO2 nanopore arrays electrode, J. Hazard. Mater. 171 (2009) 678–683. [24] J. Bai, Y. Liu, J. Li, B. Zhou, Q. Zheng, W. Cai, A novel thin-layer photoelectrocatalytic (PEC) reactor with double-faced titania nanotube arrays electrode for effective degradation of tetracycline, Appl. Catal. B Environ. 98 (2010) 154–160. [25] R. Daghrir, P. Drogui, N. Delegan, M.A. El Khakani, Electrochemical degradation of chlortetracycline using N-doped Ti/TiO 2 photoanode under sunlight irradiations, Water Res. 47 (2013) 6801–6810. [26] X. Nie, J. Chen, G. Li, H. Shi, H. Zhao, P.-K. Wong, et al., Synthesis and characterization of TiO 2 nanotube photoanode and its application in photoelectrocatalytic degradation of model environmental pharmaceuticals, J. Chem. Technol. Biotechnol. 88 (2013) 1488–1497. [27] R. Liu, C. Gao, Y.-G. Zhao, A. Wang, S. Lu, M. Wang, et al., Biological treatment of steroidal drug industrial effluent and electricity generation in the microbial fuel cells, Bioresour. Technol. 123 (2012) 86–91. [28] F. Harnisch, C. Gimkiewicz, B. Bogunovic, R. Kreuzig, U. Schröder, On the removal of sulfonamides using microbial bioelectrochemical systems, Electrochem. Commun. 26 (2013) 77–80. [29] A.M. Urtiaga, G. Pérez, R. Ibáñez, I. Ortiz, Removal of pharmaceuticals from a WWTP secondary effluent by ultrafiltration/reverse osmosis followed by

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48] [49] [50]

electrochemical oxidation of the RO concentrate, Desalination 331 (2013) 26– 34. X. Zhao, Y. Hou, H. Liu, Z. Qiang, J. Qu, Electro-oxidation of diclofenac at boron doped diamond: kinetics and mechanism, Electrochim. Acta 54 (2009) 4172– 4179. I. Sirés, J.A. Garrido, R.M. Rodríguez, P.L. Cabot, F. Centellas, C. Arias, et al., Electrochemical degradation of paracetamol from water by catalytic action of Fe2+, Cu2+, and UVA light on electrogenerated hydrogen peroxide, J. Electrochem. Soc. 153 (2006) D1–D9. I. Sirés, N. Oturan, M.A. Oturan, R.M. Rodríguez, J.A. Garrido, E. Brillas, ElectroFenton degradation of antimicrobials triclosan and triclocarban, Electrochim. Acta 52 (2007) 5493–5503. M. Skoumal, C. Arias, P.L. Cabot, F. Centellas, J.A. Garrido, R.M. Rodríguez, et al., Mineralization of the biocide chloroxylenol by electrochemical advanced oxidation processes, Chemosphere 71 (2008) 1718–1729. M. Skoumal, R.M. Rodríguez, P.L. Cabot, F. Centellas, J.A. Garrido, C. Arias, et al., Electro-Fenton, UVA photoelectro-Fenton and solar photoelectro-Fenton degradation of the drug ibuprofen in acid aqueous medium using platinum and boron-doped diamond anodes, Electrochim. Acta 54 (2009) 2077–2085. E. Guinea, J.A. Garrido, R.M. Rodríguez, P.-L. Cabot, C. Arias, F. Centellas, et al., Degradation of the fluoroquinolone enrofloxacin by electrochemical advanced oxidation processes based on hydrogen peroxide electrogeneration, Electrochim. Acta 55 (2010) 2101–2115. E. Isarain-Chávez, J.A. Garrido, R.M. Rodríguez, F. Centellas, C. Arias, P.L. Cabot, et al., Mineralization of metoprolol by electro-Fenton and photoelectro-Fenton processes, J. Phys. Chem. A 115 (2011) 1234–1242. A. Dirany, I. Sirés, N. Oturan, A. Özcan, M.A. Oturan, Electrochemical treatment of the antibiotic sulfachloropyridazine: kinetics, reaction pathways, and toxicity evolution, Environ. Sci. Technol. 46 (2012) 4074–4082. A. El-Ghenymy, R.M. Rodríguez, C. Arias, F. Centellas, J.A. Garrido, P.L. Cabot, et al., Electro-Fenton and photoelectro-Fenton degradation of the antimicrobial sulfamethazine using a boron-doped diamond anode and an air-diffusion cathode, J. Electroanal. Chem. 701 (2013) 7–13. K.P. Amorim, L.L. Romualdo, L.S. Andrade, Electrochemical degradation of sulfamethoxazole and trimethoprim at boron-doped diamond electrode: performance, kinetics and reaction pathway, Sep. Purif. Technol. 120 (2013) 319–327. A. Fabianska, A. Bialk-Bielinska, P. Stepnowski, S. Stolte, E.W. Siedlecka, Electrochemical degradation of sulfonamides at BDD electrode: kinetics, reaction pathway and eco-toxicity evaluation, J. Hazard. Mater. 280 (2014) 579–587. S.J. George, R. Gandhimathi, P.V. Nidheesh, S.T. Ramesh, Electro-Fenton oxidation of salicylic acid from aqueous solution: batch studies and degradation pathway, Clean 42 (2014) 1701–1711. M.J. Martín de Vidales, J. Robles-Molina, J.C. Domínguez-Romero, P. Cañizares, C. Sáez, A. Molina-Díaz, et al., Removal of sulfamethoxazole from waters and wastewaters by conductive-diamond electrochemical oxidation, J. Chem. Technol. Biotechnol. 87 (2012) 1441–1449. J. Wu, H. Zhang, N. Oturan, Y. Wang, L. Chen, M.A. Oturan, Application of response surface methodology to the removal of the antibiotic tetracycline by electrochemical process using carbon-felt cathode and DSA (Ti/RuO2–IrO2) anode, Chemosphere 87 (2012) 614–620. M. Rajab, G. Greco, C. Heim, B. Helmreich, T. Letzel, Serial coupling of RP and zwitterionic hydrophilic interaction LC–MS: suspects screening of diclofenac transformation products by oxidation with a boron-doped diamond electrode, J. Sep. Sci. 36 (2013) 3011–3018. E.B. Cavalcanti, S. Garcia-Segura, F. Centellas, E. Brillas, Electrochemical incineration of omeprazole in neutral aqueous medium using a platinum or boron-doped diamond anode: degradation kinetics and oxidation products, Water Res. 47 (2013) 1803–1815. H. Faber, H. Lutze, P.L. Lareo, L. Frensemeier, M. Vogel, T.C. Schmidt, et al., Liquid chromatography/mass spectrometry to study oxidative degradation of environmentally relevant pharmaceuticals by electrochemistry and ozonation, J. Chromatogr. A 1343 (2014) 152–159. M. Sh Yahya, N. Oturan, K. El Kacemi, M. El Karbane, C.T. Aravindakumar, M.A. Oturan, Oxidative degradation study on antimicrobial agent ciprofloxacin by electro-Fenton process: kinetics and oxidation products, Chemosphere 117 (2014) 447–454. The National Institute of Standards and Technology, The NIST 14 Mass Spectral Library& Search Software (NIST 2014/EPA/NIH). . WHO, Antimicrobial Resistance: Global Report on Surveillance. , (accessed 04.14). H. Baltruschat, Differential electrochemical mass spectrometry, J. Am. Soc. Mass Spectrom. 15 (2004) 1693–1706.

Please cite this article in press as: Enric Brillas, Ignasi Sirés, Electrochemical removal of pharmaceuticals from water streams: Reactivity elucidation by mass spectrometry, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.01.013